Group 1 - The core viewpoint of the report is that NVIDIA's new AI chip architecture, NVIDIA Vera Rubin, represents a comprehensive rethinking of system architecture, addressing the exponential growth in AI computing demands and redefining storage stack requirements [1][2][3] - NVIDIA's Vera Rubin platform includes six new chips designed to enhance system communication bandwidth, achieving a bandwidth that is twice that of the global internet total bandwidth, thus overcoming traditional data transfer bottlenecks [1][3] - The introduction of BlueField-4DPU allows for the management of a shared, persistent, high-speed context memory pool of up to 150TB, directly connected to GPUs, enabling dynamic allocation of up to 16TB of dedicated context space per GPU [2][3] Group 2 - The demand for NAND storage is projected to increase significantly, with an estimated 161EB needed by 2025, which would account for approximately 16% of the global NAND demand and 54% of enterprise SSD demand [4] - The Vera Rubin architecture allows for training a next-generation model with 100 trillion parameters using only a quarter of the cluster size required by the previous Blackwell system, greatly accelerating the research-to-product iteration cycle [3][4] - The AI computing throughput of a Vera Rubin data center is expected to be about 100 times that of a data center based on the Hopper architecture, significantly reducing the token generation cost for large-scale AI services to about one-tenth of the current level [3]

Core Insights - The dialogue at CES 2026 between Lenovo's CEO Yang Yuanqing and NVIDIA's CEO Jensen Huang marks a significant collaboration aimed at advancing AI technology and infrastructure [2][3][7]. Group 1: Collaboration and Innovation - Lenovo and NVIDIA announced the "Lenovo AI Cloud Super Factory," which aims to significantly reduce the "time to first token" for AI deployment and scale up to 100,000 GPUs to support trillion-parameter AI models [2][7]. - The collaboration leverages NVIDIA's latest Vera Rubin system, designed for advanced inference and AI agents, providing robust support for the initiative [2][8]. Group 2: AI Development and Strategy - Yang Yuanqing introduced the concept of "Hybrid AI," emphasizing the coexistence of personal, enterprise, and public intelligence to meet diverse needs [4][5]. - The focus on "human-centric intelligence" reflects a shift towards AI that serves individual and enterprise needs, moving beyond generic models [11][12]. Group 3: Infrastructure and Technology - The "AI Cloud Super Factory" aims to industrialize AI infrastructure, ensuring predictable performance and manageable operations, addressing the evolving demands of AI applications [8][9]. - Lenovo's liquid cooling technology will be integrated into the project, enhancing efficiency and safety in AI deployments [9]. Group 4: Market Position and Future Outlook - The partnership positions Lenovo as a leader in integrating design, manufacturing, and global deployment of AI infrastructure, while NVIDIA maintains its dominance in computational power [3][12]. - The dialogue suggests a pivotal moment in AI development, transitioning from algorithmic advancements to practical applications, with infrastructure builders like Lenovo and NVIDIA taking center stage [12].

转自：新华财经 黄仁勋认为，这一计划将消除AI部署过程中的不确定性，真正实现把AI基础设施推向产业化，实现性能可预测、部署可复现、运维可管理。英伟达加速计 算平台将为这一计划提供强大支撑，包括Blackwell Ultra GB300，以及在CES上发布的下一代训练与推理系统——NVIDIA Vera Rubin。"凭借联想在全球范 围内设计、制造、集成与部署的端到端能力，云服务提供商能够加速行动，企业也能在生产环境中真正信赖人工智能。" 杨元庆现场透露，不久前英伟达在总部为联想举办了一场"合作伙伴日"活动。双方展示了包括个人电脑、GPU服务器、数据中心、具身智能、机器人、车载 计算等成果。杨元庆认为，联想与英伟达长达30年的战略合作已迈上了一个崭新的里程碑，人工智能时代为联想和英伟达都带来了前所未有的发展机遇。 编辑：罗浩 图为会议现场（联想集团供图） 黄仁勋认为，企业实现AI落地，需要将先进模型与自身数据结合，构建专属的智能体系统。而这类系统不能仅依靠云端，它们必须部署在企业级人工智能 服务器和企业级基础设施之上，并逐步向边缘场景延伸，覆盖工厂车间、智能汽车、机器人等多个场景。"未来不仅属于云端原生智能， ...

近日2026年国际消费电子展（CES 2026）在美国拉斯维加斯启幕，联想同期举办创新科技大会 TechWorld。会上联想集团董事长兼CEO杨元庆与黄仁勋宣布最新合作"联想人工智能云超级工厂"计 划。这一计划将帮助云服务提供商极大缩短"time to first token"AI部署的时间，同时可迅速扩展规模至 十万台GPU，支持万亿参数级别的智能体和大语言模型。黄仁勋称英伟达加速计算平台将为这一计划提 供强大支撑，包括最新发布的下一代训练与推理系统——NVIDIA Vera Rubin。 ...

杨元庆现场透露，不久前英伟达在总部为联想举办了一场"合作伙伴日"活动。双方展示了包括个人电脑、GPU服务器、数据中心、具身智能、机 器人、车载计算等成果。杨元庆认为，联想与英伟达长达30年的战略合作已迈上了一个崭新的里程碑，人工智能时代为联想和英伟达都带来了前 所未有的发展机遇。 相关ETF （原标题：杨元庆：未来3-4年内联想与英伟达业务合作规模将翻四倍） 北京时间2026年1月7日，全球消费电子展(CES) 2026正式开幕，英伟达创始人兼CEO黄仁勋出席联想集团全球创新科技大会(Tech World)并发表重 磅演讲。联想集团董事长兼CEO杨元庆现场披露，联想未来3-4年内与英伟达的业务合作规模将实现翻四倍的目标。 随着人工智能的不断突破，过去一年，医疗健康、客户服务、数字化运营等众多垂直领域涌现出大量创新的智能体应用。多模态生成式人工智能 和智能体的结合，促成了AI在各行各业落地生根，蓬勃发展。黄仁勋预判，"下一个前沿趋势是将所有这些技术系统地整合进企业智能。" 黄仁勋认为，企业实现AI落地，需要将先进模型与自身数据结合，构建专属的智能体系统。而这类系统不能仅依靠云端，它们必须部署在企业级 人工智能服 ...

"人工智能云超级工厂"到底是什么？怎么用？简单解释，所谓的工厂其实相当于给用户提供的一个"算 力池套餐"，消费者购买后就能使用并扩容，可满足较为复杂的AI模型搭建需求。从目标用户来看，各 大厂商已经陆续搭建了自己的超算中心，中小厂商却可借此大大降低研发成本，从而降低"AI竞赛"入场 门槛。 （文章来源：每日经济新闻） 每经美国拉斯维加斯1月7日电（记者杨卉）当地时间1月6日，全球消费电子展CES 2026在拉斯维加斯 正式开幕。与此同时，联想集团在拉斯维加斯地标型景点Sphere举办了集团史上最大规模的年度技术创 新大会"Lenovo Tech World"。会上，英伟达创始人兼CEO黄仁勋携手联想集团董事长兼CEO杨元庆共同 发布了"联想人工智能云超级工厂"。据悉，英伟达最新发布的Vera Rubin将是该合作的重要组成部分。 杨元庆表示，这一计划将帮助云服务提供商极大缩短"time to first token"AI部署的时间，同时可迅速扩 展规模至十万枚GPU（图形处理单元），支持万亿参数级别的智能体和大语言模型。黄仁勋则强调，这 一计划将消除AI部署过程中的不确定性，真正实现把AI基础设施推向产业化， ...

专题：联想创新科技大会 新浪科技讯 1月7日上午消息，在2026年国际消费电子展（CES 2026）期间，联想创新科技大会举行， 联想集团董事长兼CEO杨元庆与英伟达创始人兼首席执行官黄仁勋宣布"联想人工智能云超级工厂"，英 伟达最新发布的Vera Rubin将是该合作的重要组成部分。 杨元庆表示，这一计划将帮助云服务提供商极大缩短"time to first token"AI部署的时间， 同时可迅速扩 展规模至十万枚GPU，支持万亿参数级别的智能体和大语言模型。 黄仁勋称，英伟达加速计算平台将为这一计划提供强大支撑，包括最新发布的下一代训练与推理系统 NVIDIA Vera Rubin。黄仁勋还表示，凭借联想在全球范围内设计、制造、集成与部署的端到端能力， 云服务提供商能够加速行动，企业也能在生产环境中真正信赖人工智能。 新浪声明：所有会议实录均为现场速记整理，未经演讲者审阅，新浪网登载此文出于传递更多信息之目 的，并不意味着赞同其观点或证实其描述。 责任编辑：王翔 专题：联想创新科技大会 新浪科技讯 1月7日上午消息，在2026年国际消费电子展（CES 2026）期间，联想创新科技大会举行， 联想集团董事 ...

记者丨孔海丽 编辑丨骆一帆 这是一场发生在拉斯维加斯，却可能影响未来科技走向的对话。 现场，联想集团董事长兼CEO杨元庆（YY）与NVIDIA创始人兼CEO黄仁勋（Jensen）展开了关于AI未 来的重磅对话，并宣布了一项重磅合作——"联想人工智能云超级工厂"。 联想集团董事长兼CEO杨元庆表示，这一计划将帮助云服务提供商极大缩短"time to first token"AI部署 的时间, 同时可迅速扩展规模至十万枚GPU，支持万亿参数级别的智能体和大语言模型。 英伟达最新发布的Vera Rubin将是该合作的重要组成部分。 黄仁勋称，英伟达加速计算平台将为这一计划提供强大支撑，包括最新发布的下一代训练与推理系统 ——NVIDIA Vera Rubin。 这不仅是两家公司的联手，也是AI真正走向"人本智能"的关键点。 "联想制造了世界上大部分的计算机。"黄仁勋表示。外界看来，杨元庆掌舵的联想，拥有着从PC到服 务器的庞大市场，是无可争议的"全球计算设备之王"；而黄仁勋麾下的NVIDIA，凭借GPU的绝对优 势，已被加冕为"全球算力之王"。当"设备之王"遇见"算力之王"，一场关于AI如何从云端走向使用的变 革也 ...

Core Viewpoint - The dialogue between Lenovo's CEO Yang Yuanqing and NVIDIA's CEO Jensen Huang at CES 2026 highlights a significant collaboration aimed at advancing AI technology through the establishment of the "Lenovo AI Cloud Super Factory," which is expected to enhance AI deployment efficiency and scalability [1][6]. Group 1: Collaboration and Technology - Lenovo and NVIDIA are collaborating to create the "Lenovo AI Cloud Super Factory," which aims to support cloud service providers by significantly reducing the "time to first token" for AI deployment and scaling up to 100,000 GPUs [1][6]. - NVIDIA's latest platform, Vera Rubin, will play a crucial role in this collaboration, providing robust support for advanced AI training and inference systems [1][6]. - The partnership is positioned as a transformative step in making AI more accessible and practical for various applications, moving from cloud-based solutions to real-world usage [2][6]. Group 2: AI Strategy and Insights - Yang Yuanqing emphasizes the need for a hybrid approach to AI, combining personal, enterprise, and public intelligence to meet diverse user needs, moving beyond generic models [4][9]. - The concept of "Hybrid AI" aims to create tailored intelligent solutions that address the specific requirements of individuals and businesses, ensuring data privacy and security [4][9]. - The collaboration seeks to address the challenges of high inference costs and data privacy concerns that have hindered the widespread adoption of AI technologies [4][10]. Group 3: Infrastructure and Future Outlook - The "Lenovo AI Cloud Super Factory" is designed to industrialize AI infrastructure, making performance predictable and deployment manageable, which is essential as AI evolves to understand complex environments [6][10]. - Lenovo's capabilities in design, manufacturing, and global deployment will complement NVIDIA's expertise in chip and acceleration platforms, enhancing the overall effectiveness of the AI solutions [7][10]. - The partnership is expected to quadruple in scale over the next two years, indicating strong growth potential in the AI infrastructure market [7].

Core Insights - The dialogue at CES 2026 between Lenovo's CEO Yang Yuanqing and NVIDIA's CEO Jensen Huang signifies a pivotal moment for the future of AI technology [3][4] - Lenovo announced a major collaboration with NVIDIA to establish the "Lenovo AI Cloud Super Factory," aimed at accelerating AI deployment for cloud service providers [3][4] Group 1: Collaboration and Technology - The partnership between Lenovo and NVIDIA is described as a convergence of the "Device King" and the "Power King," highlighting Lenovo's extensive market in computing devices and NVIDIA's dominance in GPU technology [4] - The "Lenovo AI Cloud Super Factory" aims to reduce the "time to first token" for AI deployment and scale up to 100,000 GPUs, supporting trillion-parameter AI models [3][4][9] - NVIDIA's latest platform, Vera Rubin, will play a crucial role in this collaboration, providing the necessary computational power for advanced AI applications [3][4] Group 2: AI Development and Strategy - Yang Yuanqing introduced the concept of "Hybrid AI," emphasizing the need for AI solutions that cater to individual and enterprise-specific needs rather than relying solely on generalized models [5][6][7] - The strategy aims to create personalized AI solutions, addressing concerns over data privacy and the high costs associated with generative AI [7][8] - The collaboration is positioned to transform AI infrastructure into a more industrialized and manageable resource, akin to utilities like water and electricity [14] Group 3: Future Implications - The dialogue suggests a shift in AI development focus from algorithmic advancements to practical applications, with Lenovo and NVIDIA emerging as key players in this new phase [14] - The integration of Lenovo's liquid cooling technology into the AI Cloud Super Factory plan indicates a commitment to efficient and sustainable AI infrastructure [9][13] - The partnership is expected to quadruple in scale over the next two years, reflecting the growing demand for AI capabilities [9]